CASE Network Studies and Analyses 320 - Identifying Vertical Product Differentiation in Three Polish...

3 2 0

Krzysztof Szczygielski, Iga Magda

Identifying Vertical Product Differentiation in Three PolishManufacturing Industries: an Enterprise Survey

W a r s a w , M a r c h 2 0 0 6

The materials published here have a working paper character. They may be subject to further publication. Theviews and opinions expressed here reflect the authors' point of view and not necessarily those of CASE.

The following paper is a result of the project 'Changes in Industrial Competitiveness as a Factor of Integration:Identifying Challenges of the Enlarged Single European Market' funded from the 5th Framework Programme ofthe European Community (Ref. HPSE-CT-2002-00148). The authors are solely responsible for the content ofthe paper. It does not represent the opinion of the Community and the Community is not responsible for anyuse that might be made of data appearing therein.

Keywords: product differentiation, quality, competitiveness, manufacturing.

© CASE – Center for Social and Economic Research, Warsaw 2006

Graphic Design: Agnieszka Natalia Bury

DTP: CeDeWu Sp. z o.o.

ISSN 1506-1701ISBN 97883-7178-402-6EAN 9788371784026

Publisher:

CASE – Center for Social and Economic Research

12 Sienkiewicza, 00-944 Warsaw, Poland

tel.: (48 22) 622 66 27, 828 61 33, fax: (48 22) 828 60 69

e-mail: [email protected]

http://www.case.com.pl/

2

Studies & Analyses No. 320 – Identifying Vertical Product Differentiation in Three Polish Manufacturing Industries...

Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1. Product differentiation and non-price competition in the economic theory– a short review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.1. Mainstream economics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71.2. Post-Schumpeterian approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2. Methodology of this paper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.1. Theoretical framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92.2. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3. Results of the analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.1. Identifying segments of the market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.2. Changes in segment orientation in 2002-2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173.3. Attempt to explain changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

4. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Annex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3


Krzysztof Szczygielski

Krzysztof Szczygielski studied mathematics at the Warsaw University and economics at the Warsaw School

of Economics and at the Gerhard-Mercator-University in Duisburg (Germany). Having been with CASE since

2001 his principal research interests include international economics, industrial organization and

competitiveness problems. He is currently a PhD candidate at the Warsaw School of Economics.

Iga Magda

Iga Magda graduated from the Warsaw School of Economics in 2004. She studied also at the University of

Poitiers, France. Currently she is enrolled in PhD studies at the Warsaw School of Economics and cooperates with

the CASE Foundation. She is co-author of publications on foreign trade and labour market.

4


Abstract

The paper attempts to identify indirectly vertical product differentiation in three industries of Polishmanufacturing (manufacture of glass and glass products, manufacture of other general purpose machineryand manufacture of other special purpose machinery) by examining how focus on a given group ofcustomers (segment) is related to company characteristics. Changes in companies' segment orientationbetween 2002 and 2005 are examined and the factors of these changes are discussed. The analysis is basedon a survey of 77 companies.

The data support the hypothesis that there exist segments in the consumer goods market, defined bythe income level of customers. In the capital goods market, it is shown that domestic-owned customersare the low-end segment of the market, whereas foreign-owned customers constitute a higher segment.It seems that industries producing capital goods have shifted towards higher market segments between2002 and 2005 and their principal motive was the pressure exerted by competitors, both domestic andforeign.

5


Introduction

The problem of heterogeneity of products offered on the markets has been of interest to both,the mainstream economics, which analyzed it with its Industrial Organization product differentiationmodels, and for those non-mainstream economists, who prefer the Schumpeterian approach and arereluctant to equilibrium analysis.

Regarding empirical research, apart from the literature devoted exclusively to productheterogeneity or variety1, there have been studies, which sought to apply this knowledge in theanalyses of other problems. In particular, a recent paper on the competitiveness of Polishmanufacturing industry by Wzi¹tek-Kubiak and Magda (2005) analysed, among other things, changes inthe relative unit export value2 of Polish products, which is interpreted as changes in the quality of thePolish exports to EU-15. The authors observed an increase in the RUEV of Polish aggregate exportsbetween 1996 and 2003, which suggests an increase in their quality. They also identified manufacturingbranches (as defined by the 3-digit level of the NACE-Rev-1 classification), that increased or decreasedthe quality of their products.

The first goal of this paper is to supplement these findings by a micro level analysis based on theresults of a survey conducted among 77 firms from three branches of Polish manufacturing. We willseek to identify vertical product differentiation (differences in quality) by considering the segments ofmarkets where firms operate, and to analyse the changes in their segment orientation between 2002and 2005 . Our second goal is to propose explanations to the observed changes. The small numbersof companies analysed within each industry obviously limit the generality of the conclusions, yet webelieve that this paper can at least make a small methodological contribution to the research intoquality and competitiveness problems of Polish manufacturing.

The structure of paper is as follows. In the first part, we shortly present theoretical approaches todifferentiation and product quality. In the second part, the theoretical framework of this paper andthe methodology of research is presented. In the third section we discuss the results of the analysis.Conclusions wrap up the paper.

6


1 See e.g. the book of Anderson, Thisse and de Palma (1992).2 Unit export value (UEV) is defined as the ratio of the value of (a bundle of) exported goods over their quantity measured in metric

tones. Relative unit export value was defined by Wzi¹tek-Kubiak and Magda as the proportion of the UEV of Polish exports tothe EU-15 and the UEV of the EU-15 internal exports.

1. Product differentiation and non-price competition in theeconomic theory – a short review

1.1. Mainstream economics

Mainstream economics defines product differentiation as a situation when producer introduces tothe product some characteristics that might increase the utility from this product for all theconsumers on the market (vertical product differentiation) or for some of the consumers (horizontalproduct differentiation). Historically, the theory of horizontal product differentiation is older, as it hasbeen launched by Harold Hotelling in 1926 (Hotelling 1926). However, in this paper we are going tofocus on vertical differentiation models, as they can be easily interpreted in terms of quality: higherutility of all consumers is identified with higher quality. That is why these models can be directlyapplied in the empirical analysis.

An important terminology reservation should be made at this point. Quality, in the sense of verticaldifferentiation is performance quality, i.e. the way consumer perceives the product. Otherwise, onecan frequently encounter a different approach, where high quality means low percentage ofdefectives in firm’s output, in other words a high level of the technical conformance of product andproject (conformance quality, see Reeves and Bednar 1994). This definition is often used in themanagement science3. However, in this part we will focus on performance quality while conformancequality will be briefly addressed in the empirical part only.

In the classical models of vertical differentiation (Shaked and Sutton 1983 and 1987, Gabszewiczet al. 1981) one assumes (as does Hotelling) a continuum of consumers, out of which each one hasdifferent parameters of choice. Contrary to the Hotelling model, all consumers have the samevaluation of the products, in the sense of the quality classification, but they differ in income. Moreprecisely, a certain continuous distribution of income over the population of consumers is assumed.One assumes as well, that the utility function is a product of the utility from quality of the purchasedproduct and the utility from the money that is left after the purchase.

These assumptions have several implications. First of all, in equilibrium, the price of a higher qualityproduct is higher than the price of a lower quality product. Secondly, if the number of products is finite,the consumers may be classified into “segments”, i.e. one may determine ranges of income such, thatall the consumers from a given range choose the same product. Thirdly, there exists a minimum qualitywhich will be purchased and under certain assumptions concerning the distribution of income thisquality is higher than zero. As a result, there may exist a product, which will not be chosen even ifoffered at a zero price. It can be shown that the minimum acceptable level of quality is the higher, thehigher the income of the poorest consumer is. The fourth characteristic of the model is linked to thethird one: under certain assumptions on the distribution of income or on the cost function, the numberof products which can be sold in equilibrium is limited. And finally the fifth characteristics: the numberof products which may be maintained on the market does not necessarily grow with the increase inincomes. The reason behind is that the competition among higher quality products lowers their pricesto such an extent that the product of lower quality loses all of its attractiveness. If the income spreadis limited, the products of the lowest quality will be pushed out of the market.

7


3 The topic extensively worked on in this literature is the problem of organizing the production process in a way to obtain theoptimum level of conformance quality, as well as the issue of profitability of investing in such improvement of productionprocess (e.g. Karmarkar i Pitbaldo 1997).

The second model we would like to invoke is the work of Banker et al. (1998) in which thedemand side is modelled with a representative consumer function, the number of firms ispredetermined, but the quality of the products is still endogen to the model. Each firm faces anegatively sloped demand curve, and the demand for its products is the higher, the higher is theirquality and the lower is their price, and, respectively, the lower the quality and the higher the priceset by competitors. Moreover, each firm has its own level of intrinsic demand. The choice of thequality level influences both fixed costs and variable costs. The authors have considered the case of aduopoly, in which one firm dominates (i.e. has a higher intrinsic demand) and the case of a symmetricoligopoly. The model is a two-stage game: firstly the competitors choose the level of quality and thenthe level of prices.

The authors analysed how the quality choices of the firms change, when the demand dominationof one firm over the second increases or decreases in the duopoly, or when the number of firmsgrows in oligopoly. They find that the more perfect competition may increase or decrease theaverage quality in the branch, depending on the fixed cost of quality improvement.

1.2. Post-Schumpeterian approach

Industrial Organization models are subject to criticism from the Austrian school economists, inparticular the ones following the legacy of Joseph Schumpeter. The main charge is that mainstreameconomics is unable of characterising accurately the process of competition, since it ignores“entrepreneurial rivalry” which is the essence of competition (see the review article by Stigler 1998).One of the factors allegedly not accounted for by mainstream is “entrepreneurial discovery” whichin view of the Austrians is not properly analysed by uncertainty models (Kirzner 1997). It is alsoclaimed that Industrial Organisation models are “static” (Wzi¹tek-Kubiak 2003, p.21). The quotationmarks are justified here, as some of those models certainly are not static in the mainstreameconomics sense of the word, e.g. multi-stage games. It is more about the fact that the equilibrium –a state in which no one is able to improve her position – even if exists, is by its very nature of nointerest for the researcher of competition processes. For competition implies a continuous search forimprovements, resulting in contradictory objectives and thus in a struggle, that makes it impossible tomaintain any kind of equilibrium.

Lack of formalization is undoubtedly a weakness of this approach4 and it also makes it difficult toformulate the definition of quality precisely. It seems that the technical characteristics of the product areof a lower importance here: what matters is the price realised on the market (Wzi¹tek-Kubiak, Magda2005). This is because an increase in price related to higher quality may allow the company to gainmarket shares, which is the principle goal of companies according to the Post-Schumpeterian paradigm.

8


4 “Such realistic reversion to the competitive concept of the classical economists hale not been systematically formalized intotheoretical models” (Stigler 1998, p. 535).

2. Methodology of this paper

2.1. Theoretical framework

Instead of a formal model, we will present now some theoretical assumptions, our empiricalanalysis rests on5. They are based on the mainstream models discussed above but do not seem tocontradict the Post-Schumpeterian approach either (at least some of them).

1. Quality is identified with the vertical product differentiation in the sense of higher utility to everyconsumer as in the model of Shaked and Sutton.

2. Consumers can be divided in segments, differing in their budget constraint, as in the model ofShaked and Sutton. Higher segments buy higher quality.

3. On the supply side, higher quality invokes higher costs due to more expensive and better qualityinputs, as in the model of Banker. If company is a multiproduct one, then higher average qualityimplies higher costs. We draw two consequences from this assumption:

a) Main competitors in the higher segments are companies from more developed countries

b) Bigger companies produce higher quality goods (this however applies only to sectors with highfixed costs of quality)

4. There are three reasons for which a company could be improving the quality of its products:

a) a change in the production function (e.g if company’s technology improves due to capitalaccumulation or takeover by new investors)

b) a change in the competition structure (e.g. if new low-cost competitors enter the segmentcompany currently operates in)

c) a change on the demand side6;

2.2. Methodology

The paper is based on a survey of 77 manufacturing firms from three industries: manufacture ofglass and glass products (NACE code 261), manufacture of other general purpose machinery (NACEcode 292) and manufacture of other special purpose machinery (NACE code 295). The survey wasconducted in spring 2005 and the questions referred to the situation at that time, as well as to thesituation in 2002. Some accounting data for the end of 2002 and 2004 were also asked after.

The analysis will consist of three steps:

a) Identification of market segmentations relevant for the firms;

b) Observing changes in companies’ segment orientation

c) Looking for determinants of these changes

Research methodology will be presented according to these three steps.

a) Identification of market segmentations relevant for the firmsWe hypothesise that the following kinds of segmentations of the market matter for the surveyed

companies:

9


5 We do not rule out a formalization in the next version of this paper. 6 This however might distort the old relationship between product characteristics and market segments and therefore it has some

consequences for the empirical research.

• For consumer goods: customers with an income over average, those with an average incomeand those with an income under average. The data does not allow us to distinguish domestic andforeign customers.

• For capital goods: customers-firms with a predominant share of domestic capital; customers-companies with a predominant share of foreign capital (yet based in Poland); customers-firmsbased in the EU.

The second segmentation may raise some doubts, because, strictly speaking, EU customers arenot a market segment but a different geographical market. One can identify them with a marketsegment – the highest market segment – only if two conditions are fulfilled: first, all the products weanalyse are tradable goods (so that the consumption set of the EU customers is not smaller than thatof domestic ones), and second, that they have the same valuation of quality as domestic customers.We believe that these assumptions are not too strong.

The verification of the hypotheses will consist of demand-side and supply-side analyses (in case ofthe first segmentation where only the supply-side will be examined). Consequently, we will first checkif the hypothesized segments indeed differ in terms of requirements of the customers. Companieswere asked this question directly but we will supplement their answers by examining the priceelasticity of demand in individual segments. On the supply-side we will attempt to answer thequestion if the choice of market segment was related to a certain level or quality of inputs. In case ofthe EU-customers in the second segmentation we will bear in mind that exporting requires additionalinputs such as freight costs, insurance costs, more bureaucratic work etc.

We have to make two additional remarks at this stage. First, the theoretical assumptions we outlinedabove make it necessary that we analyze each NACE industry separately: we cannot merge to one“market” products that are not substitutes or have different production functions. Unfortunately if webreak down the sample into NACE industries and we consider consumer goods and capital goodsseparately, then we end up with rather small samples of companies. This obviously limits our analysis.

Secondly, we are unable to use the one-dimensional measure of average quality of company’sproducts which would seem optimal in the light of the theoretical assumptions. This optimal measurewould be the average income of the firm’s customer. We cannot calculate that indicator, because thesegments are defined too roughly, especially in the case of capital goods. Consequently, we have noweights to calculate average quality.

What we are going to do, instead, is to look at the correlations between the share of sales realisedin each of the segments and a given characteristic of demand or supply. Since we distinguished threehypothetic segments in each of the markets, this method cannot always provide an unambiguousclassification of segments: if the correlation with the focus on segment A is significantly positive andthe correlation for B and C close to zero – then it says nothing about B and C. We believe, however,that we run a sufficiently large number of “test” so as to draw conclusions on segmentation.

b) Observing changes in companies’ segment orientationWe will observe how companies move across segments while examining the statistical significance

of the changes. Our inability to calculate the average quality of companies’ products will make thispart of the analysis particularly difficult.

c) Looking for the determinants of these changesWe will attempt to verify, among other things, two of the three hypotheses suggested by the

model of Banker (see 2.1). According to that paper improvements in quality could be (i) a result ofthe (autonomous) company development, (ii) response to competitors, (iii) consequence of changesin customer requirements. We will be able to check the two latter hypothesis by analysing thequestion after the motives for quality improvement asked in the questionnaire.

10


3. Results of the analysis

3.1. Identifying segments of the market

As explained above we will analyse separately firms manufacturing consumer goods and firmsmanufacturing capital goods and we will test separately their hypothetic market segmentations.

3.1.1. Consumer goods: segmentation by the income level of customers

Companies producing consumer goods all belong to the NACE category 261 (manufacture of glassand glass products)7. We will distinguish three groups of customers: those with an income overaverage, those with an average income and those with an income under average. We do notdistinguish between domestic and foreign customers. Since the group of companies is small (23 firms)and the empirical material proved to be rather limited, we will perform only a supply-side analysis forthis segmentation8.

Supply-side analysisWe start our research into the inputs of companies from investigating the labour factor: wages and

the level of human capital9. As evidenced by the results of the Spearman rank correlation, the higherthe share of high-income customers in company’s portfolio, the higher the average wage and therelationship is significant at the 1% level (Table 1). A reverse relationship (with high p-values) can beobserved for the middle- and low-income customers, which proves that higher wages was associatedwith the upper segment only.

The analysis of human capital shows that a higher share of high-income customers is correlatedwith a high share of marketing staff and especially – R&D staff. As expected, the correlation with theshare of low-income customers is negative (and quite strong). Interestingly, there seem to be anegative relationship between orientation on more wealthy customers and the share of workers withuniversity education (Table 2).

11


7 Three companies from the NACE group 295 (manufacture of other special purpose machinery) that declared producingconsumer goods were excluded from the analysis

8 Although in the section 3.1.2 we do perform a supply-side analysis for an even smaller groups of capital goods producers fromthe NACE industry 292, in that case we use questions about the requirements of different groups of consumers, which werenot asked consumer goods producers.

9 We have to acknowledge that several companies discussed in this section produce also capital goods, so the analysis of inputsmight be to some extent distorted by this fact.

Table 1. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. averagewage paid by the company - Spearman's correlation

High segment(customers with

over-average income)

Middle segment(customers withaverage income)

Low segment(customers with

under-average income)average wage Correlation Coefficient ,554(**) -,361 -,382

Sig. (2-tailed) ,006 ,090 ,072N 23 23 23

** Correlation is significant at the 0.01 level (2-tailed)

Regarding other inputs, companies were asked to rate the competitiveness of their owntechnology in a three-point-scale. Again, the firms which were concentrated more on the wealthycustomers thought more of its production technology, while the ones focused on the low-endsegment of the market considered their technology not competitive.

We should mention at this place, that the question if the company had an ISO certificate (or otherkinds of quality-management schemes) had no relationship whatsoever with company’s orientation interms of market segments.

To obtain further insights into the costs and quality of production processes we checked howcompanies answered the questions about the biggest cost-side obstacles to a growth in their sales.Companies were presented a list of possible obstacles and they were asked to assess each of themin a five-point-scale. The results are reported in Table 4. Little surprisingly, firms that sold primarilyto richer customers were satisfied with their technology, while the firms focused on less wealthyclients were not (the answers are strongly correlated with the answers to the previous question). Thepositive relationship between the income level of the customers and the human capital in thecompany is also confirmed.

12


Table 2. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. shares inemployment of different kinds of staff





under-average income)Pearson Correlation -,040 -,012 ,048Sig. (2-tailed) ,856 ,957 ,828Share of white-collar staffN 23 23 23Pearson Correlation -,353 ,031 ,331Sig. (2-tailed) ,107 ,893 ,133

Shareof workerswith universityeducation

N 22 22 22

Pearson Correlation ,472(*) -,089 -,408Sig. (2-tailed) ,027 ,692 ,060Share of R&D staffN 22 22 22Pearson Correlation ,357 -,010 -,345Sig. (2-tailed) ,095 ,962 ,107Share of marketing staffN 23 23 23

** Correlation is significant at the 0.01 level (2-tailed).* Correlation is significant at the 0.05 level (2-tailed).

Table 3. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. companies'opinions on the competitiveness of their technology





under-average income)Pearson Correlation ,408 ,082 -,460(*)

Sig. (2-tailed) ,054 ,711 ,027

Company’s ownrating of itstechnologyas comparedto domesticcompetitors

N 23 23 23

Pearson Correlation ,371 ,040 -,395Sig. (2-tailed) ,081 ,856 ,062

Company’s ownrating of itstechnologyas comparedto industry’s worldleaders

N 23 23 23


When comparing companies from the same industry, labour productivity, defined by revenue fromsales per employee is a relevant measure of firm’s technological and organisational efficiency. Theresults of correlation analysis show again that firms focused on higher market segments were superiorin those respects (Table 5). Interestingly, the negative correlation is significant for the middle segmentof the market.

We checked main competitors of the firms focused on different groups of customers. The resultsconfirm the difference between the high and the middle segment of the market with respect to therole of the competitors from less developed countries (Table 6).

On the other hand we found little evidence of a relationship between the size of the company andsegment orientation, apart from a notable correlation between revenue from sales and focus on thehigh segment (Table 7).

Finally, we looked at one “Schumpeterian” measure of quality competition. Companies wereasked to assess in a three-point-scale the role of quality dimension and price dimension in theircurrent “method of competition”. Intuitively, one would expect that firms focused on highersegments would attach more weight to the former method, while the ones concentrated on lower

13


Table 4. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. main cost-side obstacles to the increase in company's sales





under-average income)Pearson Correlation -,174 ,127 ,083Sig. (2-tailed) ,427 ,563 ,707Excessive labour costsN 23 23 23Pearson Correlation -,544(**) ,124 ,450(*)Sig. (2-tailed) ,009 ,584 ,036Insufficient labour skillsN 22 22 22

Pearson Correlation ,049 -,095 ,018

Sig. (2-tailed) ,829 ,673 ,935Terms of financingthe investments

N 22 22 22Pearson Correlation

-,540(**) ,157 ,424(*)

Sig. (2-tailed) ,009 ,487 ,049Outdated productiontechnology

N 22 22 22Pearson Correlation

-,207 -,122 ,290

Sig. (2-tailed) ,355 ,589 ,190Insufficient scaleof production

N 22 22 22

Pearson Correlation ,155 -,281 ,044

Sig. (2-tailed) ,492 ,205 ,844Insufficient levelor insufficient use of IT

N 22 22 22** Correlation is significant at the 0.01 level (2-tailed).* Correlation is significant at the 0.05 level (2-tailed).

Table 5. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. main cost-side obstacles to the increase in company's sales





under-average income)Correlation Coefficient

,253 -,425(*) -,037

Sig. (2-tailed) ,245 ,043 ,867

labour productivity(revenue from salesper employee)

N 23 23 23

* Correlation is significant at the 0.05 level (2-tailed).

segment would choose the latter. The signs of the correlation analysis generally support thishypothesis, yet the correlations are weak and statistically insignificant (Table 8).

3.1.2. Capital goods: segmentation by the customer company

In this section we will examine two markets for capital goods, defined by the NACE industries no.292 (manufacture of other general purpose machinery) and no. 295 (manufacture of other specialpurpose machinery). We will examine if the following groups of customer firms: domestic-ownedfirms, foreign-owned firms and EU-based firms indeed constitute different segments of the market.Our main analytical tool will be, again, the correlation analysis. To maintain the transparency of thepaper, most of the tables in this section were moved to the Annex.

14


Table 6. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. the role ofdifferent types of competitors on the domestic market





under-average income)

Pearson Correlation ,042 ,368 -,298

Sig. (2-tailed) ,853 ,092 ,178Domestic competitors

N 22 22 22

Pearson Correlation -,166 -,086 ,225

Sig. (2-tailed) ,459 ,703 ,315

Competitorsfrom the developedcountriesoutside EU N 22 22 22

Pearson Correlation -,140 -,045 ,169Sig. (2-tailed) ,534 ,842 ,452Competitors from EU-15N 22 22 22

Pearson Correlation -,317 ,412 ,028

Sig. (2-tailed) ,150 ,057 ,902Competitors from lessdeveloped countries

N 22 22 22

Table 7. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. size of thecompany – Spearman's correlation





under-average income)Correlation Coefficient ,078 ,116 -,066Sig. (2-tailed) ,724 ,598 ,765EmploymentN 23 23 23Correlation Coefficient ,349 -,240 -,204Sig. (2-tailed) ,103 ,270 ,349

Revenuefromsales asof 31.12 2004

N 23 23 23

Table 8. Percentage of sales realised in the segments defined by the income level of customer in 2005 vs. thedeclared kind of competition





under-average income)Pearson Correlation -,298 ,162 ,181Sig. (2-tailed) ,167 ,460 ,409price competitionN 23 23 23Pearson Correlation ,107 -,098 -,037Sig. (2-tailed) ,627 ,656 ,867quality competitionN 23 23 23

Demand-side analysisCompanies were asked directly if the customers with the dominant role of foreign ownership had

higher requirements with respect to the quality of the products. In the industry no. 292 a vastmajority of surveyed firms did not see any difference between the two groups of customers (Table A1, see Annex). In the branch 295, however this fraction constituted only 44% of firms, while 41%considered the foreign-owned customers to be more demanding (Table A 2).

If our segmentation of the market meets the criteria of vertical product differentiation, then highersegments of the market should have a lower price elasticity of demand. We tried to see if this indeedwas the case by analysing changes in prices and changes in the volume of sales of products that werenot changed in terms of quality (technical nor non-technical) between 2002 and 2005. Wecalculated Spearman’s correlation coefficients between shares of sales realised in each segment andthe quasi-elasticity of demand, defined by the ratio of change in sales over change in price10.

Again, in the industry 292 there seems to be no relationship between the hypothetic segments andquasi-price elasticity of demand (Table A 3). On the other hand, in the industry 295 the more were thesurveyed companies oriented at the segment of foreign-owned companies, the lower quasi-elasticity ofdemand, and the opposite is the case for the orientation on domestic-owned customers (Table A 4).

Supply-side analysisCompanies oriented more on foreign-owned companies were paying lower wages, and in case of

the industry 295 this relationship is significant (Table A 5). In the same industry the orientation on theEU-based customer-firms was correlated positively with average wage.

In the industry 292 the level of human capital is positively and quite strongly correlated with thefocus on foreign-owned customers (even if only the correlation with the share of marketing staff issignificant at 5% level). There is also a considerable negative correlation with the orientation ondomestic-owned customers (Table A 6). In case of this measure we found no significant correlationsfor the industry 295 (Table A 7). The analysis of human capital yielded no results with respect to thefocus on the EU-based customers.

Regarding opinion of firms on their technology, in both industries the companies that were sellinga larger share of their output to the EU-customers thought more on their technology and thesecorrelations are significant at the 5% level (Table A 8). It is difficult to say anything about the domesticmarket: only in case of the industry 295 we found a substantial (yet insignificant at the 5% level)negative correlation between orientation at domestic-owned customers and company’s own ratingof its technology as compared to industry’s world leaders (Table A 9).

In an attempt to assess indirectly the technological level of companies operating in the twosegments we looked again at the biggest cost-side obstacles to the increase in company’s sales. Theanalysis yielded few interesting results, however for the industry 295 the signs of the correlationcoefficients suggest that focus on foreign-owned customers was related with a higher level oftechnology while focus on domestic-owned customers – with a lower level of technology (Table A11). Note that all but one correlations with the share of foreign-owned customers are positive,which indicates that firms focused on that group of customers saw generally less cost obstacles toa growth in their sales.

We investigated the labour productivity of companies defined by the sales-employment ration. Wefound a significant negative correlation between focus on domestic-owned customers and labour

15


10 It is not the textbook price elasticity of demand, because it also reflects shifts in the demand curve. In fact for several firms thequasi-elasticity of demand we calculated was higher than unity.

productivity in the industry 295 (Table A 12). Correlation coefficients for the two remaining segmentswere positive and the one for the EU customers quite strong.

The analysis of main competitors on the domestic market suggests that in the industry 292 firmsfocused on domestic-owned customers were competing mainly with domestic companies while theones selling a larger share of their output to the foreign-owned firms customers rather withcompanies from EU-15 (Table A 13). Correlations between the role of different kinds of competitorson the domestic market and firm’s focus on the EU market are rather problematic and we will notdiscuss them. As regards the industry 295, we found no significant results.

Regarding the size of the companies, we find, again, significant correlations in the industry 292,where firms focused on foreign-owned, and especially on the EU-based customers tend to be biggerin terms of employment and revenue from sales, while the opposite is the case for the companiesconcentrated on domestic-owned customers (this holds however only for the revenue from sales,see Table A 15). Again, we found no significant correlations in the industry 295.

Finally we looked at the “Schumpeterian” measure of the role of the price and quality dimensionof competition. We found that in the industry 292 price competition was significantly and positivelycorrelated with the focus on foreign-owned customers while quality competition was significantly andnegatively correlated with the share of domestic-owned customers (Table A 17). In the industry 295we did not find any significant results apart from a quite strong negative correlation between pricecompetition and orientation on domestic-owned customers (Table A 18).

3.1.3. Conclusions from the analysis of hypothetic segmentations

Generally speaking we find that the segmentation by the income level of customers in the NACEindustry 261 (manufacture of glass and glass products), was a correct one, i.e. the groups ofcustomers we distinguished indeed constituted different segments of the market in the sense of thetheory of vertical product differentiation. Especially the difference between the high and the middlesegment seems substantial.

The results for producers of capital goods are more difficult to interpret, since different analysesyielded different results. Table 9 contains the summary of results of all “tests” we performed. Blankfields mean no conclusions or insignificant results. Question marks indicate ambiguous results.

For the industry 292 we have gained substantial evidence in support of the hypothesis thatdomestic-owned customers were the lowest segment of the market. On the other hand we cannot

16


Table 9. Summary of analyses of hypothetic segments in the capital goods marketsNo. 292 (manufacture of other general

purpose machinery)No. 295 (manufacture of other

special purpose machinery)

Rankof EU-basedcustomers

Rank of foreignowned

customers

Rankof domestic-

ownedcustomers

Rankof EU-basedcustomers

Rank of foreignowned

customers

Rankof domestic-

ownedcustomers

RequirementsDemand 2 1 3

Wage 1 3 2Human Resources 2 1 3 2 1 3

Technology 2 1 3 1 2,5 2,5Cost obstacles

Labour productivityCompetitors ? ? ?

Firm size 1 2 3 1,5 1,5 3Schumpeter 1 2,5 2,5 1 2 3

Average 1,50 1,63 2,88 1,42 1,83* 2,75* plus we know that foreign-owned customers had higher requirements with respect to quality than domestic ones in the industry 295

really tell which group of the customers constituted the highest segment: the difference between EU-based and foreign-owned customers is tiny and it might be the result not of a higher quality but of thefact that exporting products requires additional resources (correlation of EU-export intensity withthe size of the firm supports that supposition).

The situation is similar in the industry 295. Domestic-owned customers stand out as the low-endsegment. It seems that the EU-based customers are the high-end segment, yet we cannot be sureabout that.

3.2. Changes in segment orientation in 2002-2005

In this section we check if companies moved across the segment and draw conclusions aboutchanges in the quality of their products. Here we will be able to relate to the taxonomy of NACE 3-digit industries proposed in the paper of Wzi¹tek-Kubiak and Magda (2005). According to thattaxonomy the manufacture of glass and glass products (261) were included in the cluster of “winner”industries i.e. the most competitive ones and ones that had a high and growing quality of theirproducts in 1995-2003. On the other hand, both, manufacture of other general purpose machinery(292) and manufacture of other special purpose machinery (295), belonged to the export-orientedgroup, characterised by poor performance indicators and a low – but growing – quality.

3.2.1. Consumer goods

Analysis of changes in the segment orientation of the producers of consumer goods in the glassindustry (261) yields ambiguous results. On the one hand, when we look at the mean changes in theshares of sales realised in different segments, we see that both extreme segments diminished to theadvantage of the middle segment (Table 10). The point is, the decline in the upper segment is bigger,which could imply an overall decline in average quality of the output of surveyed companies (thesechanges, however, are not significantly different from zero, according to the T-test).

On the other hand, if we take a look at the number of companies that moved to higher and lowersegments we find that the former were more numerous than the latter (again, we considered that acompany moved to a higher segment if the upper segment increased more than the lower one). This resultis reported in Table 11. Apparently the upward shifts were on average smaller than the downward ones.

17


Table 10. Change in the share of sales realised in the market segments defined by the income level of the customers(in percentage points)

N Mean Std. DeviationHigh segment (customers with over-average income) 22 -1,1364 9,50336

Middle segment (customers with average income) 22 1,5909 7,77456Low segment (customers with under-average income) 22 -,4545 5,32494

Table 11. The number of companies that moved to higher/lower segments of the marketFrequency Percent

Move to a lower segment 4 18,2No change 11 50,0Move to a higher segment 7 31,8Total 22 100,0

3.2.2. Capital goods

In the companies from the industry 292 we observe a decline in the share of domestic-ownedcustomers and in the share of foreign-owned customers, and an increase in the share of EU-basedclients (Table A 19). If we were absolutely sure that clients from the EU are the high-end segment ofthe market, then these developments would have to be interpreted as an increase in the average qualityin the surveyed companies. Since we are not, we can only say that there are some arguments in favourof this statement (especially given that changes are not significantly different from zero see Table A 20)

In case of the firms from the industry 295 we can see a substantial – statistically significant – declinein the proportion of the domestic owned customers (Table A 21 and Table A 22). Yet we cannot beabsolutely sure if that implies an increase in the average quality, because of the concurrent decline inthe share of EU-based customers.

3.3. Attempt to explain changes

Firms were asked directly if they improved quality and if they did, then for what reasons. Theproblem is that virtually all the companies said they improved quality (Table 12), which wouldobviously contradict our findings – if it was not for the difference in the perception of the very notionof “quality”. In the above analyses we were referring to the performance quality. On the other hand,the firms most likely declared an improvement in the conformance quality (see discussion in 1.1).

Firms were offered several explanations why they improved the quality of their products andthese explanations refer to both performance quality and conformance quality (Table A 23).

“Strong domestic competition on the market of the products manufactured to date” stands out as themotive that played a substantial role in all three industries (note that a positive answer to this questionsuggests a change in the product characteristics so it indicates an increase in performance quality).Similarly, the import competition was important for the surveyed companies from all the industries.

Regarding the role of demand, producers of capital goods from both analysed industries (292 and295) indicated quite frequently “Growth of demand for products of higher quality in the domesticmarket”, which indicates performance quality. Note that for the producers of consumer goods in theglass industry (262) it was not among the principal reasons. Companies in all the industries attachedan importance to the “adjustment to the requirements of the customer firm” which could be anindicator of improving conformance quality (because it implies no change in firm’s target groups).

18


Table 12. Opinion of the firms on the improvements in the quality of their productsindustry Total

261 292 295Yes 24 17 28 69Did the company improve the

quality of its products No 1 1 6 8Total 25 18 34 77

4. Conclusions

In this paper we sought to identify vertical product differentiation in three industries of Polishmanufacturing. Based on an enterprise survey we analysed producers of consumer goods in theNACE industry 261 (manufacture of glass and glass products) and producers of capital goods in theNACE industries 292 (manufacture of other general purpose machinery) and 295 (manufacture ofother special purpose machinery).

In the consumer goods market we hypothesised a market segmentation defined by the incomelevel of customers. As shown by our analysis of survey data, including both supply-side and demand-side indicators, that segmentation meets very well the criteria of vertical product differentiation wediscussed in the theoretical part.

In the capital goods market we assumed a segmentation into domestic-owned customer firms,foreign-owned (but domestic-based) customer firms, and EU-based firms. The results of the “tests” weran were less straightforward than in the case of consumer goods, but we concluded that in bothindustries, 292 and 295, domestic-owned customers were the low-end segment. We were not surewhich of the remaining two groups of customers was the highest segment and which was the middle one.

Having confirmed the differences among segments we analysed changes in segment orientation ofthe surveyed companies over time, which could indicate changes in the products’ quality. We foundambiguous results in the glass industries: more companies moved up the quality ladder than down,yet the average quality could decline at the same time. For the industry 292 we found some evidencesuggesting an improvement in quality. Although we observed significant changes in the segmentorientation of the surveyed companies belonging to the industry 295, we were unable to determinewithout ambiguity that it indicated improving quality.

As for the motives behind quality improvements, companies in all the industries indicatedcompetition, both domestic and foreign as the main reason. Interestingly, changes in demand playedan important role for the producers of capital goods but less so for the producers of consumer goods.

Although the small number of surveyed companies limited both, the generality of our conclusions,and our analytical techniques, we believe that the general idea of the methodology proposed in thispaper can be used in similar studies in future.

19


References

Simon P. Anderson, Andre de Palma, Jacques-Francois Thisse 1992 Discrete Choice Theory of Product Differentiation,The MIT Press.

Rajiv D. Banker. Inder Khosla. Kingshuk K. Sinha 1998. Quality and Competition, “Management Science“, Vol. 44,No. 9 (Sep., 1998), pp. 1179-1192.

J. Jaskold Gabszewicz. Avner Shaked. John Suton. J.-F.Thisse 1981. International Trade in Differentiated Products,“International Economic Review”, Vol. 22, No. 3 (Oct., 1981), pp. 527-534.

Harold Hotelling 1929. Stability in Competiton, “The Economic Journal”, Vol. 39, No. 153 (March 1929), 41-57.

Uday S. Karmarkar. Richard C. Pitbaldo 1997 Quality, Class, and Competition, “Management Science”, Vol. 43,No. 1 (Jan., 1997), pp. 27-39.

Kirzner, I., Entrepreneurial Discovery and the Competitive Market Process: An Austrian Approach, “Journal of EconomicLiterature”, Vol. XXXV (March 1997), 60-85.

Carol A. Reeves David A. Bednar 1994. Defining Quality: Alternatives and Implications, “The Academy ofManagement Review”, Vol. 19, No. 3. Special Issue: “Total Quality” (Jul. 1994), 419-445.

Avner Shaked. John Sutton 1983. Natural Oligopolies, “Econometrica”, Vol. 51, No. 5 (Sep., 1983), 1469-1484.

Avner Shaked. John Sutton 1987. Product Differentiation and Industrial Structure, “The Journal of IndustrialEconomics”, Vol. 36, No. 2 (Dec., 1987), pp. 131-146.

George J. Stigler 1998 Competition in: John Eatwell, Murray Milgate, Peter Newman, The New Palgrave. A Dictionaryof Economics, The Macmillan Press & The Stockton Press, pp. 531-535.

Anna Wzi¹tek-Kubiak. Iga Magda Shift in quality versus changes in competitiveness. The case of Polish manufacturingin: „New World Order: Economic, Social and Political Tendencies at the Beginning of Third Millennium“,Uniwersytet £ódzki, ?ód? 2005

Anna Wzi¹tek-Kubiak Konkurencyjnoœæ polskiego przemys³u, INE PAN i Dom Wydawniczy Bellona, Warszawa2003.

20


Annex

21


Table A 1 Requirements of foreign-owned and domestic owned customers with respect to quality – industry 292

no. 292 (manufacture of other general purpose machinery)Domestic-owned customers had higher

requirements with respect to qualityYes No Same Total

Yes Amount 1 1 0 2% of Total 5,6% 5,6% ,0% 11,1%

Same Amount 0 0 16 16% of Total ,0% ,0% 88,9% 88,9%

Total Amount 1 1 16 18

Foreign-ownedcustomers had higher

requirements withrespect to quality

% of Total 5,6% 5,6% 88,9% 100,0%

Table A 2 Requirements of foreign-owned and domestic owned customers with respect to quality – industry 295

no. 295 (manufacture of other special purpose machinery)Domestic-owned customers had higher

requirements with respect to qualityYes No Same No data Total

Yes Amount 2 11 0 1 14% of Total 5,9% 32,4% ,0% 2,9% 41,2%

No Amount 1 0 0 0 1% of Total 2,9% ,0% ,0% ,0% 2,9%

Same Amount 0 0 15 0 15% of Total ,0% ,0% 44,1% ,0% 44,1%

No data Amount 1 0 0 3 4% of Total 2,9% ,0% ,0% 8,8% 11,8%

Total Amount 4 11 15 4 34

Foreign-ownedcustomers had

higherrequirements withrespect to quality

% of Total 11,8% 32,4% 44,1% 11,8% 100,0%

Table A 3 Percentage of sales realised in the segments defined by the kind of customer company vs. quasi priceelasticity of demand – Spearman's correlation analysis in industry 292

no. 292 (manufacture of othergeneral purpose machinery)

Domestic-ownedcustomers

Foreign-ownedcustomers

(based in Poland)EU-based customers

Correlation Coefficient,402 -,316 ,114

Sig. (2-tailed) ,195 ,317 ,724quasi price elasticity ofdemand

N 12 12 12

Table A 4 Percentage of sales realised in the segments defined by the kind of customer company vs. quasi priceelasticity of demand – Spearman's correlation analysis in industry 295

no. 295 (manufacture of otherspecial purpose machinery)




Correlation Coefficient,413 -,371 ,264

Sig. (2-tailed) ,079 ,118 ,275quasi price elasticityof demand

N 19 19 19

Table A 5 Percentage of sales realised in the segments defined by the kind of customer company vs. average wage– Spearman's correlation


Foreign-owned customers(based in Poland)

EU-based customers

Correlation Coefficient ,140 -,323 -,108292 average wage Sig. (2-tailed) ,591 ,205 ,671

N 17 17 18Correlation Coefficient ,187 -,370(*) ,486(**)

295 average wage Sig. (2-tailed) ,290 ,031 ,005N 34 34 32


22


Table A 6 Percentage of sales realised in the segments defined by the kind of customer company vs. shares inemployment of different kinds of staff – industry 292


Share of white-collarstaff

Share of workerswith university

educationShare of R&D staff Share of marketing

staff

Pearson Correlation -,138 -,480 -,505 -,438

Sig. (2-tailed) ,598 ,060 ,094 ,079Domestic-ownedcustomers

N 17 16 12 17Pearson Correlation ,186 ,401 ,528 ,685(**)Sig. (2-tailed) ,474 ,124 ,077 ,002

Foreign-ownedcustomers(based in Poland) N 17 16 12 17

Pearson Correlation -,103 ,192 -,006 -,179Sig. (2-tailed) ,685 ,460 ,984 ,477EU-based customersN 18 17 13 18


Table A 7 Percentage of sales realised in the segments defined by the kind of customer company vs. shares inemployment of different kinds of staff – industry 295


Share of white-collarstaff

Share of workerswith university

educationShare of R&D staff Share of marketing

staff

Pearson Correlation -,082 -,211 ,069 -,087Sig. (2-tailed) ,646 ,231 ,712 ,635Domestic-owned

customersN 34 34 31 32Pearson Correlation ,183 ,102 -,010 ,054Sig. (2-tailed) ,301 ,567 ,959 ,771

Foreign-ownedcustomersbased in Poland) N 34 34 31 32

Pearson Correlation -,172 ,171 -,024 ,182Sig. (2-tailed) ,347 ,348 ,903 ,335EU-based customersN 32 32 29 30

Table A 8 Percentage of sales realised in the segments defined by the kind of customer company vs. companies'opinions on the competitiveness of their technology – industry 292

no. 292 (manufacture of other general purpose machinery)Company’s own rating

of its technology as comparedto domestic competitors

Company’s own ratingof its technology as compared

to industry’s world leadersPearson Correlation -,316 -,240Sig. (2-tailed) ,217 ,353Domestic-owned customersN 17 17Pearson Correlation ,111 -,135Sig. (2-tailed) ,672 ,606


N 17 17Pearson Correlation ,428 ,542(*)Sig. (2-tailed) ,077 ,020EU-based customersN 18 18


Table A 9 Percentage of sales realised in the segments defined by the kind of customer company vs. companies'opinions on the competitiveness of their technology – industry 295

no. 295 (manufacture of other special purpose machinery)Company’s own rating

of its technology as comparedto domestic competitors

Company’s own ratingof its technology as compared

to industry’s world leadersPearson Correlation -,158 -,295Sig. (2-tailed) ,381 ,096Domestic-owned customersN 33 33Pearson Correlation -,159 ,066Sig. (2-tailed) ,376 ,717


N 33 33Pearson Correlation ,378(*) ,228Sig. (2-tailed) ,036 ,218EU-based customersN 31 31

23


Table A 10 Percentage of sales realised in the segments defined by the kind of customer company vs. main cost-sideobstacles to an increase in companies' sales – industry 292


Excessivelabour costs

Insufficientlabour skills

Terms offinancing theinvestments

Outdatedproductiontechnology

Insufficientscale of

production

Insufficientlevel

or insufficientuse of IT

PearsonCorrelation

,001 -,101 ,083 ,016 -,301 ,157

Sig. (2-tailed) ,996 ,700 ,752 ,953 ,241 ,548Domestic-ownedcustomers

N 17 17 17 17 17 17PearsonCorrelation

-,278 -,244 -,404 ,029 -,007 -,215

Sig. (2-tailed) ,280 ,346 ,108 ,911 ,980 ,408

Foreign-ownedcustomers(based in Poland)


,359 ,359 ,283 -,173 ,430 -,019

Sig. (2-tailed) ,143 ,143 ,254 ,492 ,075 ,940EU-basedcustomers

N 18 18 18 18 18 18

Table A 11 Percentage of sales realised in the segments defined by the kind of customer company vs. main


Excessivelabour costs

Insufficientlabour skills

Terms offinancing theinvestments

Outdatedproductiontechnology

Insufficientscale of

production

Insufficientlevel or

insufficient useof IT

PearsonCorrelation

,106 -,230 ,280 ,254 ,277 ,289

Sig. (2-tailed) ,551 ,190 ,114 ,146 ,113 ,098Domestic-ownedcustomers


-,104 -,075 -,079 -,242 -,270 -,224

Sig. (2-tailed) ,557 ,673 ,663 ,169 ,122 ,203

Foreign-ownedcustomers(based in Poland)


-,026 ,595(**) -,085 ,071 ,157 -,057

Sig. (2-tailed) ,889 ,000 ,648 ,700 ,392 ,755EU-basedcustomers

N 32 32 31 32 32 32

** Correlation is significant at the 0.01 level (2-tailed).

Table A 12 Percentage of sales realised in the segments defined by the kind of customer company vs. labourproductivity – Spearman's correlation


Foreign-ownedcustomers (based

in Poland)EU-based customers

Correlation Coefficient -,143 ,100 -,037Sig. (2-tailed) ,584 ,704 ,883292

labour productivity(revenue from salesper employee) N 17 17 18

Correlation Coefficient -,346(*) ,148 ,272Sig. (2-tailed) ,045 ,402 ,132295

labour productivity(revenue from salesper employee) N 34 34 32


24


Table A 13 Percentage of sales realised in the segments defined by the kind of customer companies vs. the role ofdifferent types of competitors on the domestic market – industry 292

no. 292 (manufacture of other general purpose machinery) Domestic-ownedcustomers

Foreign-ownedcustomers (based in

Poland)EU-based customers

Pearson Correlation ,648(**) -,459 -,487(*)Sig. (2-tailed) ,005 ,064 ,040Domestic competitorsN 17 17 18Pearson Correlation -,253 ,296 ,086Sig. (2-tailed) ,328 ,249 ,734

Competitors from thedeveloped countries outside EU

N 17 17 18Pearson Correlation -,337 ,570(*) -,134Sig. (2-tailed) ,186 ,017 ,595Competitors from EU-15N 17 17 18Pearson Correlation -,637(**) ,075 ,709(**)Sig. (2-tailed) ,008 ,783 ,001

Competitors from lessdeveloped countries


Table A 14 Percentage of sales realised in the segments defined by the kind of customer companies vs. the role ofdifferent types of competitors on the domestic market – industry 295

no. 295 (manufacture of other special purpose machinery) Domestic-ownedcustomers

Foreign-ownedcustomers (based

in Poland)EU-based customers

Pearson Correlation ,005 ,087 ,098Sig. (2-tailed) ,977 ,624 ,593Domestic competitorsN 34 34 32Pearson Correlation ,022 -,031 ,104Sig. (2-tailed) ,900 ,861 ,571

Competitors from thedeveloped countries outside EU

N 34 34 32Pearson Correlation -,203 ,209 ,057Sig. (2-tailed) ,249 ,235 ,756Competitors from EU-15N 34 34 32Pearson Correlation ,110 ,007 -,074Sig. (2-tailed) ,534 ,967 ,688

Competitors from lessdeveloped countries


Table A 15 Percentage of sales realised in the segments defined by the kind of customer companies vs. the firm size– industry 292



Foreign-ownedcustomers (based in

Poland)EU-based customers

Correlation Coefficient -,145 ,192 ,505(*)Sig. (2-tailed) ,578 ,461 ,032EmploymentN 17 17 18Correlation Coefficient -,635(**) ,627(**) ,663(**)Sig. (2-tailed) ,006 ,007 ,003

Revenue from salesas of 31.12 2004


25


Table A 16 Percentage of sales realised in the segments defined by the kind of customer companies vs. the firm size– industry 295





Correlation Coefficient ,146 -,135 ,116Sig. (2-tailed) ,410 ,446 ,526EmploymentN 34 34 32Correlation Coefficient -,169 ,000 ,269Sig. (2-tailed) ,339 ,998 ,137

Revenue from salesas of 31.12 2004

N 34 34 32

Table A 17 Percentage of sales realised in the segments defined by the kind of customer company vs. the declaredkind of competition – industry 292





Pearson Correlation -,246 ,482 -,447Sig. (2-tailed) ,341 ,050 ,063price competitionN 17 17 18Pearson Correlation -,532(*) ,379 ,181Sig. (2-tailed) ,028 ,134 ,473quality competitionN 17 17 18


Table A 18 Percentage of sales realised in the segments defined by the kind of customer company vs. the declaredkind of competition – industry 295





Pearson Correlation -,307 ,257 ,151Sig. (2-tailed) ,092 ,163 ,433price competitionN 31 31 29Pearson Correlation -,105 ,161 ,076Sig. (2-tailed) ,554 ,364 ,678quality competitionN 34 34 32

** Correlation is significant at the 0.01 level (2-tailed).

Table A 19 Change in the share of sales realised in the segments defined by the kind of customer company –industry 292

no. 292 (manufacture of other general purpose machinery) N Mean Std. Deviation Std. Error MeanChange in the share of domestic-owned customers (in p.p.) 17 -1,1176 9,91137 2,40386Change in the share of foreign-owned customers (in p.p.) 17 -,5882 8,45620 2,05093Change in the share of EU-based customers (in p.p.) 17 2,3903 8,49189 2,05959

Note: the sum of mean changes is not exactly zero, because data on shares of sales are taken from different sources: for the shares of domestic-ownedand foreign-owned customers they are based on companies' estimates. For the shares of EU-exports they are based on sales data.

Table A 20 T-test of the hypothesis that the change in the share of sales is zero – industry 292Test Value = 0


t df Sig. (2-tailed)Mean

Difference95% Confidence Interval

of the DifferenceLower Upper

Change in the share of domestic-ownedcustomers (in p.p.) -,465 16 ,648 -1,11765 -6,2136 3,9783

Change in the share of foreign-ownedcustomers (in p.p.)

-,287 16 ,778 -,58824 -4,9360 3,7595

Change in the share of EU-basedcustomers (in p.p.)

1,161 16 ,263 2,39028 -1,9758 6,7564

26


Table A 21 Change in the share of sales realised in the segments defined by the kind of customer company –industry 295

no. 295 (manufacture of other special purpose machinery) N Mean Std. Deviation Std. Error MeanChange in the share of domestic-owned customers (in p.p.) 34 -4,7941 14,69648 2,52043Change in the share of foreign-owned customers (in p.p.) 34 2,7353 11,28772 1,93583Change in the share of EU-based customers (in p.p.) 31 -2,2891 20,13168 3,61576

Note: the sum of mean changes is not exactly zero, because data on shares of sales are taken from different sources: for the shares of domestic-ownedand foreign-owned customers they are based on companies' estimates. For the shares of EU-exports they are based on sales data

Table A 22 T-test of the hypothesis that the change in the share of sales is zero – industry 295Test Value = 0

t df Sig. (2-tailed)Mean

Difference95% Confidence Interval

of the Differenceno. 295 (manufacture of other

special purpose machinery)Lower Upper

Change in the share of domestic-ownedcustomers (in p.p.)

-1,902 33 ,066 -4,79412 -9,9220 ,3337

Change in the share of foreign-ownedcustomers (in p.p.)

1,413 33 ,167 2,73529 -1,2032 6,6738

Change in the share of EU-basedcustomers (in p.p.)

-,633 30 ,531 -2,28909 -9,6734 5,0953

Table A 23 Motives behind quality improvements (firms were asked to rate each motive in a 0 to 7 scale)261 292 295 Total

N Mean N Mean N Mean N MeanReason 1:Faster growth of demand for products of higherquality in the domestic market

24 4,42 17 5,06 28 3,82 69 4,33

Reason 2:Fall in the demand of the products manufacturedto date in the domestic market

23 2,13 17 2,71 28 3,21 68 2,72

Reason 3:Strong import competition on the marketof the products manufactured to date

24 4,00 17 4,24 28 4,18 69 4,13

Reason 4:Strong domestic competition on the marketof the products manufactured to date

24 4,33 17 3,47 27 3,56 68 3,81

Reason 5Launching exports

23 3,00 16 1,44 27 2,78 66 2,53

Reason 6Small import demand for the productsmanufactured to date

23 1,65 17 1,18 26 1,73 66 1,56

Reason 7Adjustment to the requirementsof the customer firm

24 3,96 16 4,38 26 3,73 66 3,97

CASE Network Studies and Analyses 320 - Identifying Vertical Product Differentiation in Three Polish...

Economy & Finance

Transcript of CASE Network Studies and Analyses 320 - Identifying Vertical Product Differentiation in Three Polish...